Fix ComputedFieldInfo.wrapped_property pointer when a property setter is assigned

[tlambert03]

from pydantic import BaseModel, computed_field

class Thing(BaseModel):
    x: int

    @computed_field
    def y(self) -> bool: ...

    @y.setter
    def y(self, value: bool) -> None: ...

# this statement is True on `main` *only* if a setter is not used, otherwise it returns False
# this PR fixes the pointer when a setter is used.
print(Thing.model_computed_fields["y"].wrapped_property is Thing.y)

Change Summary

Fixes the pointer in ComputedFieldInfo.wrapped_property when a property setter is used

Related issue number

Checklist

• The pull request title is a good summary of the changes - it will be used in the changelog
• Unit tests for the changes exist
• Tests pass on CI
• Documentation reflects the changes where applicable
• My PR is ready to review, please add a comment including the phrase "please review" to assign reviewers

Selected Reviewer: @sydney-runkle

[codspeed-hq]

CodSpeed Performance Report

Merging #9892 will not alter performance

_{Comparing tlambert03:fix-wrapped-prop (61fd877) with main (d654a07)}

Summary

✅ 13 untouched benchmarks

[tlambert03]

please review

[sydney-runkle]

Great, thank you for fixing this @tlambert03!!

[sydney-runkle]

Waiting to merge until we get a second pair of eyes from @dmontagu, he's more familiar with the decorator logic than I am :).

[dmontagu]

I won't pretend to understand fully what's going on here, but this change seems reasonable and I can understand in principle why an update to the decorator metadata is necessary. Also the diff here leaves me feeling pretty confident that if this introduces any bugs, and that does seem unlikely, but if it does, they should be easy to address anyway. 👍